Protein identification from protein product ion spectra

Abstract

Mass spectrometry is used to identify a protein of interest. The protein is first ionized then fragmented into protein product ion. Masses of the observed product ions are compared to product ion masses calculated in silico for database protein sequences to identify product ion matches within a predetermined mass tolerance. An algorithm that weights the product ion to matches based upon one or more factors such as product ion abundance, favored cleavage sites, product ion type, precursor ion charge state and polarity is used to score the possible matches to database proteins in order to identify the protein of interest. The invention represents a “top down” approach and is particularly well-suited for identification of a protein in a complex mixture.

Description

[0001] This application claims the benefit of U.S. Provisional Application Ser. No. 60 / 382,062, filed 20 May 2002, which is incorporated herein by reference in its entirety.STATEMENT OF GOVERNMENT RIGHTS [0002] This invention was made with government support under a grant from the National Institutes of Health, Grant No. GM45372. The U.S. Government has certain rights in this invention.BACKGROUND OF THE INVENTION [0003] Over the last fifteen years, mass spectrometry has played an increasingly important role in the identification of molecules of biological interest (Mann et al., Annu. Rev. Biochem. 2001, 70, 437-473). Indeed, recent developments in mass spectrometry have been the major factors enabling proteomics (Dove, Nature Biotechnol. 1999, 17, 233-236; Blackstock et al., Trends Biotechnol. 1999, 17, 121-127; Pandey et al., Nature 2000, 405, 837-846; Anderson et al., FEBS Letters 2000, 480, 25-31). In particular, the speed, specificity, and sensitivity of mass spectrometry make i...

AI Technical Summary

Benefits of technology

[0031] Product ion masses can be experimentally determined from a product ion mass spectrum. Experimentally determined product ion masses are compared with product ion masses calculated for each member of a comparison set of database protein sequences, thereby elucidating, for each member of the comparison set, product ion matches that are within a predetermined mass tolerance. The number of product ion matches can be counted for each member of the comparison set, and matches or possible matches between the protein of interest and one or more members of the comparison set are thereby identified.

[0032] The comparison set of database proteins sequences may include all or only some of the protein sequences or subsequences included in one or more protein databases. For example, the comparison set may be limited to protein sequences or subsequences having a calculated mass that matches the mass of the protein of interest within a predetermined mass tolerance.

[0033] Optionally, product ion masses calculated for database protein sequences include product ion masses calculated for database protein sequences that have been modified to account for one or more known or predicted protein structu

Problems solved by technology

However, information present in databases is, to varying degrees, incomplete and inaccurate in relation to the mature expressed protein, due to the multitude of post-translational processing events that can occur after protein translation.

Note that protein digestion rarely provides 100% sequence coverage of the protein, due to losses of some peptides during sample handling prior to mass spectrometry, or due to the masses of the peptides falling outside the observable mass range of the instrument, therefore making complete protein characterization difficult to achieve.

Additionally, while the approach is also amenable to the analysis of simple protein mixtures, provided that sufficient peptide masses are obtained to unambiguously identify each component of the mixture, peptide mass fingerprinting is generally not suited to the analysis of peptides resulting from proteolysis of complex protein mixtures, as the presence of peptides from many different proteins makes it difficult to assign individual peptides to their correct proteins.

Importantly, this method is error tolerant as one, or several, of the regions of the sequence tag (the tag itself or the flanking mass regions) may contain errors due to post-translational modifications yet still result in an unambiguous assignment.

However, as the sequence tag approach requires some user intervention prior to database analysis, it has not generally been employed for large scale, high throughput applications.

Hence, the approach is generally considered to be more labor intensive and expensive in terms of both time and sample consumption than the database searching approaches described above.

Unfortunately, several classes of proteins, notably hydrophobic proteins, low abundance proteins, and those with extremes of isoelectric point (pI) and molecular weight are poorly represented in 2D-gel based separations (Gygi et al., Proc. Natl. Acad. Sci. U.S.A.

However, digestion of an unfractionated protein mixture greatly increases the number of components to be analyzed and condenses the resultant peptide mixture into a narrow mass range, thereby complicating the task of isolating individual components for further analysis, placing greater demands on the performance of the mass spectrometer.

Furthermore, a problem common to all peptide sequencing approaches is that MS / MS spectra often yield insufficient product ions, product ions corresponding to cleavages not included in the search algorithms, or lack product ions with sufficient signal-to-noise, to allow their identification (Simpson et al., Electrophoresis 2000, 21, 1707-1732).

Finally, it is common that many of the peptides resulting from digestion are not observed, making complete characterization of the protein difficult to achieve.

While protein mass is fundamentally informative, it alone is not particularly useful in unambiguously identifying a protein due to the potential for post-translational processing to cause differences between the predicted and experimentally observed masses.

A major issue associated with implementation of the top-down approach on most types of tandem mass spectrometers is that the spectra derived from the dissociation of multiply-charged proteins ions typically contain product ions with charge states ranging from +1 up to the charge of the precursor ion, thereby creating possible ambiguities in assigning product ion mass and charge.

A limitation of the “double isolation” approach is that the ions from a given protein charge state are distributed over several charge states during the ion / ion reaction, thereby diluting the protein ion signal and decreasing the sensitivity for subsequent isolation and dissociation.

Although techniques that increase “charge state purification” such as double isolation of the precursor ion and ion parking have made the “top down” approach to mass spectrometry-driven protein sequencing more reliable, difficulties in matching product ion spectra with protein database information still frequently arise, especially when the information in databases is incomplete and inaccurate in relation to the mature expressed protein.

The identification and characterization of proteins present in complex mixtures remains an important analytical problem.

Images